Pika Labs’ recently unveiled AI video creator tool, Pika 1.0, is now open to everyone through Discord as well as a web version. The web version offers a variety of extra features such as using images for prompts, images and texts together as prompts, or even videos together with images and text.
A quick and creative way to use Pika Labs is to create an image through an AI model such as OpenAI’s DALL-E 3 and then using Pika to bring that image to life with animations. This process involves uploading the static image and dictating the preferred animation style through a brief textual description.
Pika Labs further enhances this experience by offering customizable features such as varying camera angles, adaptable aspect ratios, and seamless integration of video with textual elements. Moreover, users have the flexibility to modify specific segments of the produced video and adjust the backdrop scaling to suit different formats, like transforming a square 1:1 ratio into a widescreen 16:9.
The standard video output ranges from 8 to 24 frames per second and initially spans four seconds. However, with just a couple of clicks, these videos can be lengthened in four-second increments, up to a maximum duration of 15 seconds. It’s worth noting that as the duration of a scene increases, maintaining consistency becomes more challenging, as illustrated in the example below.
A more effective strategy is to generate multiple brief, high-quality scenes and then amalgamate them into an extended video through a video editing software, such as Adobe’s Premier Pro or Davinci Resolve. This method was employed to produce the showcased video using Pika Labs’ tools.
Pika’s web version can be accessed here. This webpage shows many impressive videos created through the Pika Labs community as well as the prompts used for these videos so it can easily be replicated.
Pika Labs, a rising star in the video AI startup sector and on par with industry peer RunwayML in terms of funding, is still in the early stages of its development. The company has successfully secured a significant investment of $55 million through initial funding stages, including pre-seed and seed rounds spearheaded by notable figures Nat Friedman and Daniel Gross, followed by a Series A funding round driven by Lightspeed Venture Partners.